INDEX

Explanations

terms related to physical objects and their organization

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 conhecer

-0.41

 kautta

-0.39

InputModule

-0.38

역

-0.37

 charmed

-0.37

 keskus

-0.37

PostConstruct

-0.37

 virtudes

-0.36

 פני

-0.36

ุม

-0.36

POSITIVE LOGITS

isContained

0.58

 يتيمه

0.56

 HttpNotFound

0.55

lotl

0.43

WebElementEntity

0.43

MethodManager

0.42

 saites

0.42

 Prot

0.42

 TextAlign

0.41

Зноскі

0.41

Activations Density 0.015%

terms related to physical objects and their organization

No Comments

No Known Activations